对抗样本三元组约束的度量学习算法

针对已有三元组约束的度量学习算法大多利用先验知识构建约束,一定程度上制约了度量学习算法性能的问题,本文借鉴对抗训练中样本扰动的思想,在原始样本附近学习对抗样本以构造对抗三元组约束,基于对抗三元组和原始三元组约束构建了度量学习模型,提出了对抗样本三元组约束的度量学习算法(metric learning algorithm with adversarial sample triples constraints,ASTCML)。实验结果表明,提出的算法既克服了已有固定约束方法受先验知识影响大的问题,也提高了分类精度,说明区分更加难以区分的三元组约束能够提升算法的性能。     

Compositional metric learning for multi-label classification

Multi-label classification aims to assign a set of proper labels for each instance, where distance metric learning can help improve the generalization ability of instance-based multi-label classification models. Existing multi-label metric learning techniques work by utilizing pairwise constraints to enforce that examples with similar label assignments should have close distance in the embedded feature space. In this paper, a novel distance metric learning approach for multi-label classification is proposed by modeling structural interactions between instance space and label space. On one hand, compositional distance metric is employed which adopts the representation of a weighted sum of rank-1 PSD matrices based on component bases. On the other hand, compositional weights are optimized by exploiting triplet similarity constraints derived from both instance and label spaces. Due to the compositional nature of employed distance metric, the resulting problem admits quadratic programming formulation with linear optimization complexity w.r.t. the number of training examples.We also derive the generalization bound for the proposed approach based on algorithmic robustness analysis of the compositional metric. Extensive experiments on sixteen benchmark data sets clearly validate the usefulness of compositional metric in yielding effective distance metric for multi-label classification.     

基于图像分层树的图像语义分割方法

针对传统语义分割模型缺乏空间结构信息,无法准确地描述对象轮廓的问题,提出了一种基于图像分层树的图像语义分割方法。分层树模型采用结构森林方法生成轮廓模型,为防止过度分割,运用超度量轮廓图算法得到多尺度轮廓图,然后利用支持向量机训练多尺度轮廓图生成图像分层树,通过随机森林精炼分层树,最终输出图像语义分割结果。在测试实验中,像素精确度达到82.1%,相比区域选择方法(Selecting Regions)提升了2.7%。并在较难区分的树和山脉的预测精确度上,相比层次标记方法(Stacked Labeling)分别提升了16%,25%,具有更高的稳定性。实验结果表明,在复杂的室外环境下,对图像语义分割的精确度、稳定性和速率均有明显改善。     

智能投币机在城市公交上使用的改进分析

通过鱼骨图的方法对现有智能公交投币机存在识别率偏低的问题进行了分析,提出在智能投币机中采用闭式五角叶轮或增加硬币预处理部分,对现有的智能公交投币机进行改进,使智能公交投币机的硬币辨识率得到提高.     

数控机床的可靠性分析

从不同侧面分析了数控机床的可靠性,阐述了提高数控机床可靠性的方法。     

自动焊机监测与参数化曲线控制系统设计

设计了自动焊机的监测与参数化曲线控制系统,以可视化方式为工艺人员改变焊接过程的控制参数提供了便利,该控制系统的数据库的数据存储与查询功能为工艺人员进行离线工艺分析提供了大量的数据,且该系统与焊机的数控系统密切配合,实现了焊接自动生产线的可靠监控.     

CO2气体保护焊机操作注意事项

介绍了焊工在CO2气体保护焊机的安装、使用、维护保养及维修中的注意事项。     

机械零件刚度的模糊可靠性设计

介绍了机械零件刚度模糊可靠性设计变形分布参数和刚度隶属函数的确定,导出了零件刚度模糊可靠度的计算公式,并通过机床主轴模糊可靠性设计实例说明机械零件模糊可靠性的计算方法.     

数控机床位置伺服在线调试与动态性能分析研究

介绍一种在数控机床位置伺服系统在线调试中应用系统动态性能分析方法的基本原理，给出了一种位置伺服系统动态响应曲线、动态性能分析及其主要技术指标的实例。     

New method for determination of the pose of the grinding wheel for thinning drill points

Traditionally, twist drills are reconditioned by thinning the web so the correct chisel edge length is restored. Recently, thinning has been included in the original design of drills so as to reduce torque and tool force. Because the International Standards Organization (ISO) has a system which can comprehensively model conventional twist drills but cannot model thinning specifications, this paper presents a system for precise mathematical modeling and CNC control of a 6-axis grinding workstation for drill thinning. The presented method determines the position and orientation of the grinding wheel based on the evaluated rake and clearance angles of ISO standards for 2-flute twist drills. The mathematical model and background are discussed. For verification and demonstration, two experimental drills are produced to the identical ISO standard except that one is thinned. The modeling herein is of value to industry and research if incorporated into computer software for drill design and manufacture. It is suitable for linear notch-type cutting with controlled variable rake angle along the secondary cutting edge for purposes of thinning, notching, dubbing and advanced drill research.